There are total of 54 distinct models, which can be organized into 4 groups:
The following layout helps understand how each of the models was constructed. The columns count the number of terms on the first level, the rows shows what predictors are added to the second level. Replace the stars in the name of the model with F for models with only the fixed effects and R1, R2, R3, and R4 for models with the corresponding number of random terms in the second level.